|Publication number||US7034736 B1|
|Application number||US 10/980,051|
|Publication date||Apr 25, 2006|
|Filing date||Nov 2, 2004|
|Priority date||Nov 2, 2004|
|Publication number||10980051, 980051, US 7034736 B1, US 7034736B1, US-B1-7034736, US7034736 B1, US7034736B1|
|Inventors||Ahmed Mohamed Abdelatty Ali|
|Original Assignee||Analog Devices, Inc.|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (17), Non-Patent Citations (4), Referenced by (8), Classifications (8), Legal Events (5)|
|External Links: USPTO, USPTO Assignment, Espacenet|
1. Field of the Invention
The present invention relates generally to differential processing systems.
2. Description of the Related Art
A variety of modern signal processing systems (e.g., analog-to-digital converters) process differential signals along differential paths with converter output signals obtained as the difference between the paths' differential output signals. This differential processing has historically been effective in reducing even-order harmonic energy. As the speed and linearity demands of these processing systems relentlessly increase, however, it is now often found that signal harmonics remain excessive.
Embodiments of the present invention are directed to differential processing systems and methods that reduce even-order harmonic energy. The novel features of these embodiments are set forth with particularity in the appended claims. These embodiments will be best understood from the following description when read in conjunction with the accompanying drawings.
Embodiments of the present invention are directed to the recognition that it is difficult (if not impossible) to perfectly match transfer functions of the paths of differential processing systems and that it is also difficult to generate perfectly matched differential signals so that some degree of signal imbalance will always exist in these systems. In response to these recognitions, processing system embodiments are provided which are particularly suited to reduce even-order harmonic energy generated by path and signal imbalances. These embodiments also reduce offset errors which are a form of even-order harmonic energy. Included are embodiments that address signal imbalances so that they also reduce DC input signals which are another form of even-order harmonic energy.
In the first processing mode, first and second signal portions Vinp and Vinn of a differential input signal 32 are processed along the first and second network paths 30 and 31 as indicated by the processing paths 34 and, in the second processing mode, respectively processed along the second and first network paths as indicated by the processing paths 35 (wherein subscripts p and n indicate positive and negative signal portions).
The coupling circuit 24 includes an input switch circuit 36 and an output switch circuit 37 which are configured to steer the signal portions along the paths described above (the switches are all shown in the first processing mode in
Imbalances in the first and second network paths (i.e., differences in path transfer functions) will generate even-order harmonics when first and second signal portions of a differential signal are respectively processed over those paths. Because the system 20 selectively directs each of the first and second signal portions over different ones of the first and second network paths, the effects of these path imbalances are substantially reduced with a consequent enhancement of processing linearity and a lowering of even-order harmonics in the differential output signal 33.
If the selection processes of
The signals processed through the differential processing system 20 may be analog, digital or mixtures thereof. In an exemplary processing system, the network paths 30 and 31 may convert a differential analog input signal 32 to a corresponding digital output signal.
The system further includes the mode-selection generator 26 of
The inverter 46 of
As shown in
In a first operational mode, the switches 56 are momentarily closed and the switches 53 and 54 are momentarily placed in contact with their switch nodes A to thereby capture samples of the input signal on the capacitors 51 and 52. The switches 56 are then opened and the switches 53 and 54 are moved from their switch nodes A to complete the capture and to permit the captured signal sample to be transferred to processing systems via differential port 57. This capture process is repeated successively as long as the sampler is in its first operational mode.
In a second operational mode, the switches 56 are momentarily closed and the switches 53 and 54 are momentarily placed in contact with their switch nodes B to thereby capture samples of the input signal on the capacitors 51 and 52 in a reversed orientation. The switches 56 are then opened and the switches 53 and 54 are moved from their switch nodes B to complete the capture and to permit the captured signal sample to be transferred to processing systems via differential port 57. This capture process is repeated successively as long as the sampler is in its second operational mode. After capture of the signal samples, they are passed through the output port 57 to initiate the conversion processes of the differential converter stages (44 in
The first and second operational modes can be selected so that signal samples are directed over different signal paths. Accordingly, the distorting effects of path imbalances are substantially reduced with a consequent enhancement of processing linearity and a lowering of even-order harmonics in the differential output signal. If the selection is random, the even-order harmonic energy is converted to random noise. Preferably, the downstream switches 56 close and open just prior to the closing and opening of the upstream switches 53 and 54 to thereby reduce the capture of switching-transient energy in the capacitors.
The upstream isolation provided by the downstream current gain (upstream current loss) of the buffers 62 and 63 has been found to enhance the accuracy of the signal samples. This isolation is further enhanced by inserting third and fourth buffer transistors 66 and 67 ahead of the first and second switches 53 and 54 (each buffer has a corresponding current source 68).
A bootstrap system is formed by bootstrap modules 71 that are each coupled between source and gate of a corresponding one of the switch transistors.
A second switch 75 couples the bottom plate BP to the output of the upstream buffer 66 and a third switch 76 couples the top plate TP to the gate of one of the switch transistors(53A, 53B, 54A and 54B of
When the mode command is in a first mode, a first clock state closes switch 74 to insert a bias current in the bootstrap capacitor 72 via diode 73 and a supply voltage (e.g., Vdd). For simplicity of description, it is assumed there are no voltage drops across the diode 73 and the switch 74 so that a potential of substantially Vdd is established across the bootstrap capacitor. The switch 77 is closed in this clock state so that the switch transistor 53/54 is off.
When the clock transitions to a second state, switches 74 and 77 are opened and switches 75 and 76 are closed. A signal sample Ssmpl is thus coupled from the upstream buffer 66 to the bottom plate BP. Accordingly, the top plate TP applies a bias of Ssmpl+Vdd to the gate of the switch transistor 53/54 so that a constant gate-to-source bias of Vdd is established across this transistor as indicated in
Because the bias of the switch transistor includes the signal sample Ssmpl, a constant bias voltage is applied to the switch transistor 53/54 so that its on resistance is substantially constant and the signal sample is transferred to the downstream buffer (62 or 63 in
When the mode command transitions to a second mode, the multiplexer 78 sets the switch commands 79 to thereby lock the switches 74, 75, 76 and 77 in the positions shown in
When the bootstrap module 71 is used to form the bootstrap modules in
The first and second mode commands thus route the first and second signal portions of the differential signal 32 over different signal paths. The first and second signal portions respectively pass through buffers 62 and 68 in one mode and respectively pass through buffers 63 and 62 in another mode. The distorting effects of path imbalances are thus reduced, processing linearity is enhanced and even-order harmonics are lowered.
In a first processing mode of the system 80, first and second signal portions Vinp and Vinn of the differential signal are respectively processed along the first and second processing paths 82 and 83. In a second processing mode, first and second inverted versions of the first and second signal portions are formed by the inverters 84 and 85 and these inverted versions are processed along the first and second signal paths after which the corresponding signal outputs of the first and second processing paths are inverted in the inverter 86 (the inverter is not activated in the first processing mode). Similar to the processing system 20 of
This method of reducing harmonic energy not only addresses path imbalances (imbalances in the first and second processing paths) but also signal imbalances (imbalances in the differential nature of the first and second signal portions Vinp and Vinn of the differential input signal 32). The system selectively (e.g., randomly) processes these signal portions and the inverted versions of these signal portions and, when it processes the inverted versions, it then inverts the system's output signals.
The invertible samplers 102 and 103 provide differential samples of the first and second signal portions in the first processing mode and invert differential samples in the second processing mode. The converter stages 104 then successively convert the differential samples to corresponding digital signals at the output port 106 and the mode-selection generator 26 selectively commands the first and second processing modes.
The sampling switches 112 and 114 are then returned to the open positions shown and transfer switches 115 are closed to transfer the captured signal charge into output capacitors 118 that provide feedback about a differential amplifier 119 which drives an output port 120. The differential output signals of the amplifier are then differentially and successively converted to digital signals in succeeding converter stages (104 of
The sampling operations during the second processing mode of the invertible sampler 110 of
Between these operations, the reversing switches 93 and 94 are switched to positions opposite those shown in
The mode-selection generator 26 of
As mentioned above, embodiments of the invention address even-order harmonic energy generated by system path imbalances. The embodiments are effective for reducing even-order harmonics in the system output signal by, for example, converting them into random noise. It is further noted that the energy portion converted can be controlled by selecting the ratio of times devoted to the first and second processing modes.
In the system 20 of
Accordingly, the conversion in specific processing systems can be controlled to optimize a tradeoff between harmonic performance and noise performance. It is important to note that offset errors are a form of even-order harmonic energy which are included in the system path imbalances. Therefore, embodiments of the invention also reduce offset errors.
Other embodiments of the invention address even-order harmonic energy generated by both system path imbalances and input signal imbalances. For example, the differential processing system 100 of
In an application of these embodiments, performance of a differential processing system can be measured with and without the correction processes (i.e., without use of the second processing path 35 of
Simulations were run on system embodiments of the invention. A first simulation was directed to an exemplary differential system that had no input signal imbalance but a 10% path imbalance. When operated single-endedly, the system exhibited a 65 dB second harmonic (that is, the second harmonic was 65 dB below the fundamental signal). When operated differentially, the system exhibited an 85 dB second harmonic and 110 dB third harmonic. With use of the differential processing of embodiments of the invention, the second harmonics were virtually eliminated, the third harmonic remained at 110 dB and, although the noise level was raised, the signal-to-noise ratio remained at 87 dB.
A second simulation was directed to an exemplary differential system that had a 5% input signal imbalance and a 1% path imbalance. When operated single-endedly, the system exhibited a 65 dB second harmonic. When operated differentially, the system exhibited an 85 dB second harmonic and a 110 dB third harmonic. With use of the differential processing of embodiments of the invention that did not address signal imbalance (e.g., the systems 20 and 40 of
With use of the differential processing of embodiments of the invention that addressed signal imbalance (e.g., the systems 80 and 100 of
The teachings of embodiments of the invention can be applied to various differential processing systems such as the exemplary pipelined analog-to-digital converter (ADC) 140 of
Each pipelined stage 144 converts a respective analog signal to that stage's predetermined number of digital bits and passes an amplified residue signal Sres to a succeeding converter stage 144. As each succeeding stage converts its received residue signal, its preceding stage is converting a succeeding analog input signal. All converter stages, therefore, are simultaneously converting succeeding analog input signals to their respective digital bits with final converted words issuing from an associated control and correction logic 146 at the same rate as the sampling rate in the sampler 142.
Broken expansion lines 150 in
Generally, one or more redundant bits are generated in the converter stages and the control and correction logic 146 includes circuits (e.g., full adders) that use the bits of succeeding stages to correct preceding-stage errors that result from various degrading effects (e.g., offset and/or gain errors) and also includes circuits (e.g., shift registers) that time-align the corresponding digital bits.
In addition to the sampler 142, additional samplers may be positioned between converter stages to temporarily hold the residue signals and, thereby, facilitate conversion processing. The enhanced linearity of the samplers of embodiments of the invention significantly improves the accuracy of the samples presented to the conversion stages of the ADC 140 and thereby significantly improves the performance of the ADC.
Although the invertible sampler 110 of
The embodiments of the invention described herein are exemplary and numerous modifications, variations and rearrangements can be readily envisioned to achieve substantially equivalent results, all of which are intended to be embraced within the spirit and scope of the invention as defined in the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4896156 *||Oct 3, 1988||Jan 23, 1990||General Electric Company||Switched-capacitance coupling networks for differential-input amplifiers, not requiring balanced input signals|
|US5001725 *||May 19, 1989||Mar 19, 1991||Teknekron Communications Systems, Inc.||Differential switched-capacitor sigma-delta modulator|
|US5030954 *||Sep 17, 1990||Jul 9, 1991||General Electric Company||Double rate oversampled interpolative modulators for analog-to-digital conversion|
|US5594445 *||Nov 2, 1993||Jan 14, 1997||Vlsi Technology, Inc.||Pipelined analog to digital converters and interstage amplifiers for such converters|
|US5844431||Sep 18, 1996||Dec 1, 1998||Exar Corporation||Low noise low power CMOS correlated double sampler|
|US5945872||Nov 6, 1997||Aug 31, 1999||Analog Devices, Inc.||Two-phase boosted CMOS switch drive technique and circuit|
|US6060937||Sep 9, 1999||May 9, 2000||Analog Devices, Inc.||Two-phase bootstrapped CMOS switch drive technique and circuit|
|US6072355 *||Jan 22, 1998||Jun 6, 2000||Burr-Brown Corporation||Bootstrapped CMOS sample and hold circuitry and method|
|US6255865||Nov 3, 1999||Jul 3, 2001||Nanopower Technologies Inc.||Track-and-hold circuit|
|US6323697||Jun 22, 2000||Nov 27, 2001||Texas Instruments Incorporated||Low distortion sample and hold circuit|
|US6400302||Feb 26, 2001||Jun 4, 2002||Analog Devices, Inc.||Quasi-differential successive-approximation structures and methods for converting analog signals into corresponding digital signals|
|US6404376 *||Jun 30, 2000||Jun 11, 2002||Texas Instruments Incorporated||Capacitor array having reduced voltage coefficient induced non-linearities|
|US6518901||Jun 18, 2001||Feb 11, 2003||Stmicroelectronics S.R.L.||Boosted switch device for a sampler of an analog/digital converter, and operating method thereof|
|US6525682||May 3, 2001||Feb 25, 2003||Hrl Laboratories, Llc||Photonic parallel analog-to-digital converter|
|US6650263||Nov 4, 2002||Nov 18, 2003||Analog Devices, Inc.||Differential sampler structures with reduced distortion and current demand|
|US6686860 *||Dec 12, 2000||Feb 3, 2004||Massachusetts Institute Of Technology||Reconfigurable analog-to-digital converter|
|US6724236||Oct 12, 2002||Apr 20, 2004||Texas Instruments Incorporated||Buffered bootstrapped input switch with cancelled charge sharing for use in high performance sample and hold switched capacitor circuits|
|1||Brannon, Brad, "Overcoming Converter Nonlinearities with Dither", Analog Devices Application Note AN-410, Norwood, MA, pp. 1-8, no date.|
|2||Jakonia, Darius, "RF Sampling Receiver", Department of Electrical Engineering, Linkoping University, Linkoping, Sweden, pp. 1-4, no date.|
|3||Karki, Jim, "Fully Differential Amplifiers Remove Noise from Common-Mode Signals" EDN Magazine, Nov. 9, 2000, pp. 149-156.|
|4||Kikkert, C. J., et al., "Reducing Distortion in Microwave Analogue to Digital Conversion", Electrical and Computer Engineering, James Cook University, Townsville, Australia, pp. 1-4, no date.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7436344 *||Jul 24, 2007||Oct 14, 2008||Matsushita Electric Industrial Co., Ltd.||Pipeline A/D converter|
|US7479909 *||May 20, 2005||Jan 20, 2009||Infineon Technologies Ag||Suppression of even-numbered harmonic distortion in a non-linear element|
|US7719452 *||Sep 23, 2008||May 18, 2010||Analog Devices, Inc.||Pipelined converter systems with enhanced linearity|
|US7940091 *||Mar 2, 2009||May 10, 2011||Linear Technology Corporation||Bootstrapped switch for sampling voltages beyond the supply voltage|
|US8723707||Jul 27, 2012||May 13, 2014||Analog Devices, Inc.||Correlation-based background calibration for reducing inter-stage gain error and non-linearity in pipelined analog-to-digital converters|
|US20070296623 *||Jul 24, 2007||Dec 27, 2007||Matsushita Electric Industrial Co., Ltd.||Pipeline a/d converter|
|US20100073210 *||Sep 23, 2008||Mar 25, 2010||Analog Devices, Inc.||Pipelined converter systems with enhanced linearity|
|CN103703685A *||Jun 7, 2012||Apr 2, 2014||密克罗奇普技术公司||Distributed bootstrap switch|
|U.S. Classification||341/162, 327/91, 341/122, 327/337|
|Cooperative Classification||H03M1/168, H03M1/0614|
|Nov 2, 2004||AS||Assignment|
Owner name: ANALOG DEVICES, INC., MASSACHUSETTS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:ALI, AHMED MOHAMED ABDELATTY;REEL/FRAME:015952/0668
Effective date: 20041029
|Jun 5, 2007||CC||Certificate of correction|
|Oct 26, 2009||FPAY||Fee payment|
Year of fee payment: 4
|Sep 25, 2013||FPAY||Fee payment|
Year of fee payment: 8
|Sep 25, 2017||MAFP|
Free format text: PAYMENT OF MAINTENANCE FEE, 12TH YEAR, LARGE ENTITY (ORIGINAL EVENT CODE: M1553)
Year of fee payment: 12